神经形态工程学
计算机科学
计算机体系结构
记忆电阻器
终端(电信)
晶体管
电子工程
人工神经网络
人工智能
工程类
电气工程
电压
计算机网络
作者
Li Zhu,Jia Cai,Zheng Yu Ren,Wen Xiong,Qing Wan
标识
DOI:10.1002/9783527835317.ch3
摘要
One of the main tasks in neuromorphic systems would be the design of neuromorphic devices, capable of mimicking multiple fundamental synaptic computation behaviors. Moreover, mimicking dendrite integration behaviors on a single neuromorphic device would be highly suitable for neuromorphic engineering. Recently, the study on solid-state neuromorphic devices leads to the emergence of a new branch of neuromorphic engineering owing to its great potentials. With the developments of new materials technology and new conceptual devices, several kinds of neuromorphic devices have been proposed, including two terminal resistance switch devices and three terminal transistors. Moreover, memtransistors have been reported with interesting neuromorphic functions. Furthermore, multi-electrodes have been integrated within these devices, demonstrating advanced neural functions. Especially, multi-gates could be assembled in a single transistor, exhibiting interesting dendritic integration functions and dendritic algorithm. Moreover, important neural cognitive behaviors have been mimicked on these multiterminal devices. All these results indicate their great potentials in neuromorphic engineering.
科研通智能强力驱动
Strongly Powered by AbleSci AI